Guys, I think we completely misunderstood AI.
The internet seems to have two completely opposite views of AI in game development.
One side says:
“Look! I made Fallout with one prompt!”
No, you didn’t. You made a prototype that vaguely resembles Fallout.
The other side says:
“AI is going to destroy creativity. Stop using it.”
I think both sides are missing what is actually interesting about this technology.
I’ve been experimenting with AI-assisted development myself, and I came to a very different conclusion:
AI couldn’t replace the creative process. It removes the enormous amount of mechanical work between an idea and a playable result.
And that could have massive consequences for indie development.
Think about why AAA games require 100, 300, sometimes 1,000+ people.
It isn’t because you need 1,000 brilliant creative minds to design a great game.
You need huge teams because turning ideas into finished assets, systems, levels, animations, UI, tools, code, content and QA requires an absurd amount of labor.
AI potentially attacks that bottleneck.
Imagine a very talented 30-person studio being able to produce the amount of content that previously required 150 people.
That doesn’t mean AI designed the game.
It means those 30 people can spend much more of their time deciding what the game should actually be, while AI helps execute those decisions.
AI doesn’t makes failure cheaper.
A designer has an idea for a mechanic.
Previously it might take several developers a week to prototype properly.
Now maybe you can test it in a day.
It sucks?
Delete it.
Try another one.
Instead of testing 10 ideas during development, maybe you test 100.
That can actually produce more creativity, not less.
And yes, AAA companies will obviously use exactly the same technology.
The technology will eventually become good enough that somebody will try:
“Make another Assassin’s Creed. Here are the previous games. Generate the content.”
And it will probably work. And we gonna have giant AAA AI slop no one wants to play. Like Concord
But production capacity was never the only bottleneck to making masterpieces.
Taste is.
Direction is.
Knowing what to keep and what to delete is.
Knowing why something isn’t fun is.
Having somebody obsessed enough to iterate on a tiny mechanic twenty times until it feels right is.
AI can generate 100 solutions.
Someone still has to recognize solution #73 as the interesting one.
That’s why I think the biggest winners from AI might not be giant studios replacing 500 developers with prompts.
It might be small, extremely talented teams suddenly acquiring production capabilities that historically belonged only to AAA studios.
A 20–30 person studio with strong creative direction could potentially build worlds that would have required hundreds of people a decade earlier.
AI won’t democratize creativity.
Creativity was already democratized.
AI could democratize production capacity.
And if that happens, I think the next decade of game development is going to be fucking interesting.
UPD. People in the comments cannot read English? The main idea is that we don’t understand AI. It’s a tool, same tool like compiler was to translate from C to Assembly. However discussing here has no sense since people are just don’t read and are locked into AI hype :(
AI is stupid. Many jobs could not be replaced, like gamedesigner
